This subproject is one of many research subprojects utilizing the resources provided by a Center grant funded by NIH/NCRR. The subproject and investigator (PI) may have received primary funding from another NIH source, and thus could be represented in other CRISP entries. The institution listed is for the Center, which is not necessarily the institution for the investigator. The key objective of this computational project is to rationally design small-molecule antagonists that block the interaction between the human CD4 cell-surface receptor and the gp120 envelop protein of HIV-1 as potential therapeutics for the treatment of AIDS. The viral envelop protein, gp120, undergoes a large conformational change upon binding to the cellular CD4 receptor allowing subsequent binding to the chemokine receptor and viral-host cell fusion1. The NIH PO1 GM 56550 project team (Structure-Based Antagonism of HIV-1 Envelope Function in Cell Entry) has synthesized and assayed a series of NBD compound analogs 2. These compounds compete with CD4 binding to gp12 and induces structuring of gp120 in a manner similar to CD4 binding3. A predicted binding mode for this class of compounds has been produced from computational docking studies and verified with mutational analysis4 . Over the course of the start-up current allocation AMBER 9.09. based molecular dynamics on PSC's Big Ben cluster has been used to explore the dynamic fluctuations of the inhibitor NBD556 bound to wild type and various mutations of the gp120 envelope. Analysis of the molecular dynamics trajectories indicates an asymmetric protein-ligand interactions of the tetramethyl piperidine groups. This suggests that the tetramethyl groups can be synthetically modified without negative impact to binding affinity. The trajectories of the wild-type and mutant proteins-ligand complexes are also being used to predict binding affinity using the MM/GBSA9, 10 methodology as implemented in AMBER. These calculations are currently underway using the start-up allocation which expires in April, 2009. The current MAC proposal requests computational resources on Pople to continue molecular dynamics and MM/GBSA calculations using the latest version (10) of AMBER with a recently solved crystal structure. The request also includes resource for installation of the ROCS shape-based virtual screening software6-8 for discovery of novel compounds. A medium allocation will provide resource for the five computational steps necessary to complete a compound design cycle for novel NBD compounds. These calculations are: 1) Produce a correlation plot of predicted verses measured binding affinities using AMBER/MM-GBSA for existing compounds 2) Generate correlations between calculated and experimental binding affinities of compound and mutant-gp120 pairs 3) Analyze protein-ligand interactions of compound and mutant-gp120 pairs to ascertain which regions of gp120 and the small molecule are likely to enhance binding affinity 4) Identify novel compounds using the ROCS based similarity search of the Zinc11 database that encapsulate new chemotypes and interactions 5) Use the AMBER/MM-GBSA methodology to predict binding affinity of novel inhibitors prior to synthesis. Rapid turn-over using the Pople version of AMBER 10.0 for molecular dynamics with ROCS virtual screening would greatly enhance the structure-based design process of HIV entry inhibitors.